Value of multi-b value DWI in the assessment of early cerebral changes in asymptomatic HIV-positive adolescents

Abstract

Background

Magnetic resonance imaging (MRI) and functional MRI techniques have been widely used in the diagnosis of human immunodeficiency virus (HIV) infection related diseases.

Purpose

To explore whether magnetic resonance diffusion-weighted imaging (DWI) can track water molecular diffusion changes in the brain of asymptomatic HIV-positive adolescents.

Material and Methods

Multi-b value DWI was performed in 23 adolescents, including 15 HIV-positive participants and eight HIV-negative healthy participants. Mean apparent diffusion coefficient (ADC), slow apparent diffusion coefficient (ADCs) values, fast apparent diffusion coefficient (ADCf) values, distribution diffusion coefficient (DDC) values, and heterogeneity index (α) values were calculated within regions of interest (ROIs) in the frontal lobes, basal ganglia, and temporal lobe. Non-parametric tests were then performed.

Results

In the bilateral frontal lobes, the mean α values in HIV-positive participants were significantly increased compared with those in healthy participants (right side P = 0.001; left side P = 0.000). In the left frontal lobe, the mean DDC value in HIV-positive participants was significantly increased compared with that in healthy participants (P = 0.008). In the bilateral frontal lobes, the mean ADCf values in HIV-positive participants were significantly lower than those in healthy participants (right side P = 0.011; left side P = 0.008). In the left basal ganglia, the mean α values in HIV-positive participants were significantly lower than that in healthy participants (P = 0.013).

Conclusion

Multi-b value DWI could reflect the early characteristics of water molecule diffusion in HIV infections.

Keywords

Asymptomatic HIV-positive brain images central nervous system early cerebral changes multi-b value diffusion-weighted imaging (DWI)

Introduction

Approximately 36.9 million people worldwide were living with HIV in 2014. In China, more than 0.57 million people have been infected with HIV. Over the past 15 years, the number of young people dying of acquired immune deficiency syndrome (AIDS) in the world has increased twofold. The majority of AIDS adolescents live in developing countries (1). AIDS severely harms adolescents’ health. Therefore, adolescents with AIDS have become a focal point in AIDS research.

HIV spreads widely throughout the body soon after infection. Although viral invasion of the brain occurs in the early stage, probable changes during this period are not well characterized (2). In such cases, the focus of change is not signal intensity in the brain, but heterogeneity to a minor extent (3).

Advanced magnetic resonance imaging (MRI) is an attractive imaging modality in clinical and research studies due to its non-invasive nature, excellent soft tissue contrast, provision of both functional and anatomical information, and lack of ionizing radiation (4). MRI and functional MRI techniques have been a widely applied application in the diagnosis of HIV infection related diseases (5 –11). Diffusion-weighted imaging (DWI) can reconstruct the physiological structure and pathological state of in vivo tissue according to the Brownian motion of water molecules (12,13). Conventional DWI, which is typically performed using two b-values, is widely used in the diagnosis, prognosis, and treatment assessment of diseases (12 –15). Water diffusion is generally quantified by a mono-exponential DWI model. This model shows that the diffusion mode of water molecules within a voxel is single. Then the diffusion rate within a voxel can be quantified by means of the ADC value. However, in the brain, DW signal decay is multi-exponential (16,17). In the bi-exponential model, water molecules diffusion includes two modes, fast and slow diffusion, which may well describe the multi-exponential signal decays (16,17). To provide more extensive information about water molecules diffusion in the brain, Bennett developed the stretched-exponential DWI method. The stretched-exponential model introduced a new parameter (α), that varies between 0 and 1. α is defined as the heterogeneity index, which is related to the degree of intravoxel water diffusion heterogeneity (18,19). According to the stretched-exponential model, there is a continuous distribution rate within a voxel. The distribution diffusion coefficient (DDC) can be regarded as an apparent diffusion coefficient (ADC) approximation that is weighted by the volume fraction of water in each part of the continuous distribution (4).

Recently, research in HIV-positive patients using conventional DWI (8,9,11) has made some progress. However, multi b-value DWI in the study of HIV-positive patients is less reported. Furthermore, the previous studies did not evaluate the differences between asymptomatic HIV-positive participants and HIV-negative healthy participants. Asymptomatic HIV-infected adolescents, a group that is often neglected, is hardly taken into consideration in scientific research (5,8,20).

Some scholars indicated that the central nervous system is involved in the early stage of HIV infection (5,8,20). The aim of our study was to explore whether DWI, by analyzing the parameters of DWI, can track water molecular diffusion changes in the brain of early stage of asymptomatic HIV-positive adolescents.

Material and Methods

Patients

Twenty-three adolescents (13 men, 10 women; mean age, 16.4 ± 1.4 years; age range, 14–18 years) were included. Fifteen adolescents (8 men, 7 women; median age, 17 years) were born with HIV due to mother-to-child transmission. The remaining eight adolescents (5 men, 3 women; median age, 17 years) did not have HIV and were regarded as healthy control participants. A CD4T cell count and HIV RNA examination were performed, and the results were confirmed by the State Confirmation Laboratory. In this study, exclusion criteria were as follows: (i) history of neurological diseases (e.g. opportunistic intracerebral infections and brain neoplasms); (ii) history of psychiatric disease (e.g. depression and schizophrenia); and (iii) definite changes in brain structure and the presence of abnormal signal on conventional MRI examination.

All of the participants and data were obtained from Zhongnan Hospital of Wuhan University. The study was approved by the School of Physical Science and Technology, Wuhan University, and the ethics committee of the Zhongnan Hospital of Wuhan University. All participants and their parents were informed of the scanning process before the study and submitted their written approval letters to participate in the study.

MRI

All participants were examined with a 3.0T MR scanner (Siemens Magnetom Trio 3.0T, Erlangen, Germany). Whole-brain conventional MRI involved the following parameters: sagittal TSE sequence T1-weighted (T1W) imaging (TR, 700 ms; TE, 9.6 ms); axial TSE-FS sequence T2-weighted (T2W) imaging (TR, 2000 ms; TE, 76 ms); and coronal TSE-TIRM (turbo inversion recovery magnitude) sequence T2W imaging (TR, 3800 ms; TE, 68 ms). Multi-b value DWI was performed with an EPI sequence. The following scanning parameters were employed: TR, 3400 ms; TE, 85 ms; flip angle (FA), 90°; number of excitations, 4; matrix, 192 × 192; scanning slices, 20; slice thickness, 5 mm; inter-slice distance, 0 mm; and field of view, 219 × 219 mm. The scan time was 62–65 s (Figs. 1 –3).

Fig. 1.

Images from an HIV-positive 18-year-old man. DWIs with b-values of (a) 0 s/mm²; (b) 50 s/mm²; (c) 150 s/mm²; (d) 200 s/mm²; (e) 400 s/mm²; (f) 600 s/mm²; and (g) 800 s/mm² in the frontal lobe. The red ellipse in the DWIs is ROIs.

Fig. 2.

Images from an HIV positive 16-year-old girl. DWIs with b-values of (a) 0 s/mm²; (b) 50 s/mm²; (c) 150 s/mm²; (d) 200 s/mm²; (e) 400 s/mm²; (f) 600 s/mm²; and (g) 800 s/mm² in the basal ganglia. The red ellipse in the DWIs is ROIs.

Imaging processing

Statistical parameter mapping (SPM) software was used to realign all multi-b value DW images and perform space smoothing. Oval regions of interest (ROIs) were placed in the bilateral frontal lobe, bilateral basal ganglia, and bilateral temporal lobe on DWIs with MRIcro Software. The sizes of the ROIs were in the range of 95–100 mm². The average signal strength was obtained from each ROI separately. The DWI signal strength was fitted with MATLAB R2014a software. Then ADC values, ADCs values, ADCf values, DDCvalues, and α values were calculated. ADCf values represent incoherent microcirculation within the voxel (perfusion-related diffusion or the fast component of diffusion) and ADCs represent pure molecular diffusion (slow component of diffusion) (21). Three DWI models fit the mean signal intensity for b = 0, 50, 150, 200, 400, 600, and 800 s/mm² images using the following equations.

The mono-exponential model is mathematically described as follows:

S (b) / S_{0} = \exp (- ADC \times b)

where S(b) is the signal intensity with DW value, S0 is the signal intensity when b = 0 s/mm², b is the b-value, which is calculated for a standard square-shaped gradient pulse pair as follows:

b = γ 2 \times G 2 \times δ 2 (Δ - δ / 3)

Here, γ is the gyromagnetic ratio, G is the strength of the motion probing gradients (MPGs), δ is the duration of one MPG pulse, and Δ is the interval between the leading edges of the MPG pulses. ADC is the apparent diffusion coefficient (7).

The bi-exponential model is mathematically described as follows:

S (b) / S_{0} = f \exp (- {ADC}_{f} \times b) + (1 - f) \exp (- {ADC}_{s} \times b)

where S(b), S0, and b are defined as above. f is the fraction of the diffusion related to microcirculation. ADCs is the slow apparent diffusion coefficient and ADCf is the fast apparent diffusion coefficient.

The stretched-exponential model is mathematically described as follows:

S (b) / S_{0} = \exp {- (DDC \times b) a}

where S(b), S0, and b are defined as above. DDC is the distribution diffusion coefficient and α is the heterogeneity index.

Statistical analysis

Statistical analysis was performed with SPSS Statistics 19.0 software (SPSS Inc., Chicago, IL, USA). Differences between the HIV-positive participants and healthy control participants were assessed with independent samples non-parametric tests according to each DWI parameter (ADC, ADCs, ADCf, DDC, and α). P values < 0.05 were considered statistically significant. Receiver operating characteristic (ROC) curve analysis was performed using SPSS 19.0. The diagnostic accuracy was calculated by measuring the area under the ROC curve (Az) (Figs. 4 –6).

Fig. 3.

Images from an HIV-positive 15-year-old boy. DWIs with b-values of (a) 0 s/mm²; (b) 50 s/mm²; (c) 150 s/mm²; (d) 200 s/mm²; (e) 400 s/mm²; (f) 600 s/mm²; and (g) 800 s/mm² in the temporal lobe. The red ellipse in the DWIs is ROIs.

Fig. 4.

The αROC and DDC ROC curve in the frontal lobe.

Fig. 5.

The ADCf ROC curve in the frontal lobe.

Fig. 6.

The αROC curve in the left basal ganglia.

Results

There were no definite changes in brain structure or abnormal signals on conventional MRI examination.

The results are displayed in Tables 1 –3. In the bilateral frontal lobe, the mean α value in HIV-positive participants (right α = 0.941 ± 0.034; left α = 0.931 ± 0.029) was significantly increased compared with that in healthy participants (right α = 0.861 ± 0.050; left α =0.842 ± 0.049; P = 0.001 and P = 0.000). In the left frontal lobe, the mean DDC value in HIV-positive participants (DDC = 0.866 ± 0.052 × 10^–3mm²/s) was significantly increased compared with that in healthy participants (DDC = 0.792 ± 0.055 × 10^–3mm²/s, P =0.008). In the bilateral frontal lobe, the mean ADCf value in HIV-positive participants (right ADCf =0.256 ± 0.136 × 10^–2mm²/s; left ADCf = 0.424 ±0.224 × 10^–2mm²/s) was significantly reduced compared with that in healthy participants (right ADCf =1.417 ± 1.907 × 0^–2mm²/s; left ADCf = 1.630 ±1.588 × 10^–2mm²/s; P = 0.011 and P = 0.008).

Table 1.

Mean α, DDC, and ADC values in ROIS (^–x ± s).

Participants	ADC	ADCs	ADCf	α	DDC
R HIV + FL	0.835 ± 0.059	0.663 ± 0.209	0.256 ± 0.136	0.941 ± 0.034	0.85 ± 0.055
L HIV + FL	0.837 ± 0.038	0.650 ± 0.133	0.424 ± 0.224	0.931 ± 0.029	0.866 ± 0.052
R Healthy FL	0.819 ± 0.034	0.746 ± 0.052	1.417 ± 1.907	0.861 ± 0.050	0.798 ± 0.216
L Healthy FL	0.816 ± 0.028	0.723 ± 0.113	1.630 ± 1.588	0.842 ± 0.049	0.792 ± 0.055
R HIV + BG	0.782 ± 0.070	0.683 ± 0.134	1.172 ± 1.822	0.906 ± 0.067	0.730 ± 0.104
L HIV + BG	0.805 ± 0.078	0.618 ± 0.205	0.811 ± 0.966	0.848 ± 0.068	0.707 ± 0.092
R Healthy BG	0.819 ± 0.088	0.720 ± 0.125	0.338 ± 0.334	0.931 ± 0.036	0.782 ± 0.094
L Healthy BG	0.793 ± 0.067	0.628 ± 0.214	0.784 ± 0.622	0.923 ± 0.050	0.756 ± 0.073
R HIV + TL	1.020 ± 0.354	0.733 ± 0.177	1.608 ± 2.059	0.834 ± 0.102	1.032 ± 0.568
L HIV + TL	1.084 ± 0.492	0.805 ± 0.183	2.037 ± 1.821	0.826 ± 0.104	0.960 ± 0.463
R Healthy TL	0.887 ± 0.073	0.799 ± 0.181	1.579 ± 1.386	0.861 ± 0.100	0.883 ± 0.093
L Healthy TL	0.816 ± 0.028	0.723 ± 0.113	1.630 ± 1.588	0.842 ± 0.049	0.792 ± 0.151

The units of ADC, ADCs, and DDC are 10^–3mm²/s. The unit of ADCf is 10^–2mm²/s. The unit of α is 1.

BG, basal ganglia; FL, frontal lobe; HIV + , HIV-positive; L, left; L Healthy FL, the values in the left frontal lobe of healthy participants; R, right; R HIV + FL, the values in the left frontal lobe of HIV-positive participants; TL, temporal lobe.

In the left basal ganglia, the mean α values in HIV-positive participants (α = 0.848 ± 0.068) were significantly lower than those in healthy participants (α = 0.923 ± 0.050; P = 0.013). No significant difference in α values existed in the right basal ganglia (P > 0.05). In addition, the results also did not reveal any significant differences in the ADC, ADCs, ADCf, and DDC values (all P > 0.05) in the bilateral basal ganglia.

There was no significant difference in any of these parameters between the two groups in the bilateral temporal lobe (Tables 1 and 2).

Table 2.

Non-parametric tests P values.

P value	ADC	ADCs	ADCf	α	DDC
Right FL	0.392	0.548	0.011	0001	0. 087
Left FL	0.238	0.115	0.008	0.000	0.008
Right BG	0.169	0.428	0.325	0.548	0.265
Left BG	0.506	0.875	0.591	0.013	0.294
Right TL	0.776	0.825	0.825	0.428	0. 149
Left TL	0.076	0.925	0.506	0.466	0.975

BG, basal ganglia; FL, frontal lobe; TL, temporal lobe.

Table 3.

The area under the ROC curve.

Participants	The area under the ROC curve
ADC_f ROC in the left FL	0.833
ADCf ROC in the right FL	0.825
αROC in the left FL	0.983
αROC in the right FL	0.900
DDC ROC in the left FL	0.833
αROC in the left BG	0.817

BG, basal ganglia; FL, frontal lobe.

The area under the αROC curve (Az) was 0.817 in the left basal ganglia. The areas under the αROC curve, DDC ROC curve, and ADCf ROC curve were 0.983, 0.833, and 0.833, respectively, in the left frontal lobe. The areas under the αROC curve and ADCf ROC curve were 0.900 and 0.825, respectively, in the right frontal lobe (Table 3).

Discussion

To study characteristics of water molecular diffusion in the early stage of HIV infection, a group of asymptomatic HIV-positive adolescents was used. By analyzing the parameters of DWI, the change in brain water molecule diffusion in HIV infection was tracked. In this study, the results showed significant differences in α, ADCf, and DDC values between asymptomatic HIV-positive participants and healthy control participants. These findings indicated that multi-b value DWI could reflect HIV preclinical abnormalities that conventional imaging modalities are unable to reflect.

In the frontal lobes in the stretched-exponential model, the mean α values in the bilateral frontal lobes of HIV-positive participants were significantly increased compared with those of healthy participants. In the bi-exponential model, the mean ADCf values in the bilateral frontal lobes of HIV-positive participants were significantly reduced compared with those in healthy participants. Cloak et al. reported that the ADC values of HIV-infected individuals were significantly increased compared with those of healthy controls in the both frontal lobes (22). López et al. reported abnormalities in the frontal lobes in AIDS dementia complex patients shown by magnetic resonance spectroscopy (MRS) (23). Gongvatana et al. (10) and Pomara et al. (24) found that the fraction anisotropy (FA) values in the frontal lobes identified with diffusion tensor imaging (DTI) in HIV-infected individuals differed significantly from those in healthy controls. Both these results and our result indicate significant differences in the frontal lobes.

The ADCf values differed significantly between the two groups, but no significant difference was found in ADC values. In the study, none of the HIV-infected individuals had mental illness, opportunistic brain infections, or intracranial tumors. In addition, no definite changes in brain structure or abnormal signals were noted on conventional MRI examination. One hypothesis is that the brain was chronically infected, resulting in cortical atrophy and decreased cell water content. At the same time, the density of the tissue increased, so there was no significant change in the cell gaps. Another hypothesis is that the ADCf value is more greatly affected by blood circulation and microcirculation perfusion. The ADCf value is approximately one order of magnitude higher than the ADCs value.

In the basal ganglia, the results showed abnormalities in the unilateral basal ganglia of HIV-infected participants. In the left basal ganglia, the mean α values in HIV-positive participants were significantly reduced compared with those in healthy participants (P = 0.013). This finding indicates that the water molecular diffusion is abnormal in the brain of HIV-positive patients. Bekiesinska et al. made a summary of unilateral basal ganglia lesions in children and adults (25). Bladowska et al. found that the relative blood volume values in HIV-infected individuals were significantly lower than those in healthy controls on the right side of the temporo-parietal cortex and the left side of the prefrontal cortex (5). In the study, none of the HIV-infected individuals had mental illness, opportunistic brain infections, or intracranial tumors. In addition, no definite changes in brain structure or abnormal signals were noted on conventional MRI examination. These researchers thought that perfusion-weighted imaging (PWI) could reflect changes in unilateral brain regions before HIV induces neurocognitive disorders. The scenarios in asymptomatic HIV-positive participants in our study were similar to those in Bladowska’s study. These results also revealed abnormalities in unilateral brain regions of HIV-infected participants.

Many studies demonstrate abnormalities in the bilateral basal ganglia of some patients (26 –28). Corti et al. found abnormalities in the bilateral basal ganglia of AIDS patients with central nervous system cryptococcosis (26). Hegde et al. showed bilateral basal ganglia lesions in a 37-year-old HIV-seropositive man with CNS toxoplasmosis by MRI (27). Lee et al. noted abnormalities in the bilateral basal ganglia of a diabetic uremic patient by DWI (28). In fact, there are no contradictions between these studies despite the different research objectives There were opportunistic infections, brain tumors, or disease infection in their participants, whereas our participants were asymptomatic HIV-positive individuals.

We did not identify any significant differences in the temporal lobes. Gongvatana et al. showed abnormalities in both temporal lobes in hepatitis C coinfection with HIV with DTI (10). Sheerin et al. observed a high cortical signal in the right temporal lobe in a Herpes simplex encephalitis patient with DWI (29). However, all of these patients showed clinical symptoms.

In this study, water molecule diffusion in the brains of asymptomatic HIV-positive participants was abnormal. Although they were in the asymptomatic stage, these findings may still reflect structural or functional brain damage in the early stage of clinical symptom development. This finding is consistent with the results of previous studies (5,6,30). Bladowska et al. suggested that PWI could reflect changes in brain regions before HIV induces neurocognitive disorders (5). Mueller-Mang et al. studied the cervical spinal cord in asymptomatic HIV-positive patients by DTI. They found mean FA values and mean diffusivity (MD) values that exhibited significant differences between asymptomatic HIV-positive patients and controls (6). Ances et al. examined the impact of HIV on rCBF within the BG and visual cortex in 33 HIV-1-positive neurologically unimpaired participants and reported significantly reduced rCBF within the BG and visual cortex in both acute/early and chronic HIV-infected patients (30). They concluded that perfusion disorders, which occur soon after seroconversion, may reflect neuronal or vascular injury of the brain in HIV-infected individuals before they even exhibit opportunistic infections (30). This study also showed that before the appearance of clinical symptoms and imaging features of opportunistic intracerebral infections, α values in the basal ganglia and α, DDC, and ADCf values in the frontal lobes were abnormal. These abnormal values could reflect water molecular diffusion abnormalities in the brains of HIV-positive individuals, which may be attributed to the fact that the corresponding brain area has been damaged. Of course these abnormalities are the result of many factors, and the dominant factor remains uncharacterized.

There are some limitations in this research. First, we only investigated the basal ganglia, frontal lobe, and temporal lobe. More regions will be explored in future research. Second, because the participants in this study were asymptomatic HIV-positive adolescents, many difficulties were experienced in collecting these samples, resulting in a small sample size.

In conclusion, both the bi-exponential and stretched-exponential models could provide more extensive information than the mono-exponential model and better describe the diffusion heterogeneity of water molecules within a small range in the brain of HIV-infected patients. Both of the two models reflect the early brain characteristics of HIV-infected individuals and may serve as a reference for early changes in the central nervous system in HIV infections.

Footnotes

Acknowledgments

The authors thank the Medical Imaging Division of the Zhongnan Hospital of Wuhan University.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Natural Science Foundation of China (10875092), the Natural Science Foundation of Hubei Province, China (2012FKB04449), and the Elekta (Shanghai) Instruments Ltd Teaching and Research Foundation.

References

Radfar

Sedaghat

Banihashemi

. Behaviors influencing human immunodeficiency virus transmission in the context of positive prevention among people living with HIV/acquired immunodeficiency syndrome in Iran: A qualitative study. Int J Prev Med 2014; 5: 976–983.

Davis

Hjelle

Miller

. Early viral brain invasion in iatrogenic human immunodeficiency virus infection. Neurology 1992; 42: 1736–1739.

Ragin

Ochs

. Structural brain alterations can be detected early in HIV infection. Neurology 2012; 79: 2328–2334.

Mukherji

Chenevert

Castillo

. Diffusion-weighted magnetic resonance imaging. J Neuroophthalmol 2002; 22: 118–122.

Bladowska

Knysz

Zimny

. Value of perfusion-weighted MR imaging in the assessment of early cerebral alterations in neurologically asymptomatic HIV-1-positive and HCV-positive patients. PLoS One 2014; 9: e102214–e102214.

Mueller-Mang

Law

Mang

. Diffusion tensor MR imaging (DTI) metrics in the cervical spinal cord in asymptomatic HIV-positive patients. Neuroradiology 2011; 53: 585–592.

Harezlak

Cohen

Gongvatana

. Predictors of CNS injury as measured by proton magnetic resonance spectroscopy in the setting of chronic HIV infection and CART. J Neurovirol 2014; 20: 294–303.

Zacharia

Law

Naidich

. Central nervous system lymphoma characterization by diffusion-weighted imaging and MR spectroscopy. J Neuroimaging 2008; 18: 411–417.

Soto-Hernández

Moreno-Andrade

Góngora-Rivera

. Nocardia abscess during treatment of brain toxoplasmosis in a patient with aids, utility of proton MR spectroscopy and diffusion-weighted imaging in diagnosis. Clin Neurol Neurosurg 2006; 108: 493–498.

10.

Gongvatana

Cohen

Correia

. Clinical contributors to cerebral white matter integrity in HIV-infected individuals. J Neurovirol 2011; 17: 477–486.

11.

Park

Lee

. Progressive multifocal leucoencephalopathy isolated to posterior fossa in a patient with AIDS: DWI and 1H-MRS features. Korean Society of Radiology 2010; 63: 403–407.

12.

Boyer

Gonçalves

Lee

. Magnetic resonance diffusion-weighted imaging: reproducibility of regional apparent diffusion coefficients for the normal fetal brain. Ultrasound Obstet Gynecol 2013; 41: 190–197.

13.

Partridge

McDonald

. Diffusion weighted magnetic resonance imaging of the breast: protocol optimization, interpretation, and clinical applications. Magn Reson Imaging Clin N Am 2013; 21: 601–624.

14.

Baumann

Cammoun

Conus

. High b-value diffusion-weighted imaging: A sensitive method to reveal white matter differences in schizophrenia. Psychiatry Res 2012; 201: 144–151.

15.

Kwee

Galban

Tsien

. Intravoxel water diffusion heterogeneity imaging of human high-grade gliomas. NMR in Biomedicine 2010; 23: 179–187.

16.

Maier

Bogner

Bajzik

. Normal brain and brain tumor: multicomponent apparent diffusion coefficient line scan imaging. Radiology 2001; 219: 842–849.

17.

Lee

Springer

Jr . Effects of equilibrium exchange on diffusionweighted NMR signals: the diffusigraphic ‘shutter-speed’. Magn Reson Med 2003; 49: 450–458.

18.

Bennett

Schmainda

Rowe

. Characterization of continuously distributed cortical water diffusion rates with a stretched-exponential model. Magn Reson Med 2003; 50: 727–734.

19.

Bennett

Hyde

Rand

. Intravoxel distribution of DWI decay rates reveals C6 glioma invasion in rat brain. Magn Reson Med 2004; 52: 994–1004.

20.

Lentz

Kim

Lee

. Changes in MRS neuronal markers and T cell phenotypes observed during early HIV infection. Neurology 2009; 72: 1465–1472.

21.

Bihan

Turner

MacFall

. Effects of intravoxel incoherent motions (IVIM) in steady? state free precession (SSFP) imaging: application to molecular diffusion imaging. Magn Reson Med 1989; 10: 324–337.

22.

Cloak

Chang

Ernst

. Increased frontal white matter diffusion is associated with glial metabolites and psychomotor slowing in HIV. J Neuroimmunol 2004; 157: 147–152.

23.

López-Villegas

Lenkinski

Frank

. Biochemical changes in the frontal lobe of HIV-infected individuals detected by magnetic resonance spectroscopy. Proc Natl Acad Sci USA 1997; 94: 9854–9859.

24.

Pomara

Crandall

Choi

. White matter abnormalities in HIV-1 infection: a diffusion tensor imaging study. Psychiatry Res 2001; 106: 15–24.

25.

Bekiesinska-Figatowska

Mierzewska

Jurkiewicz

. Basal ganglia lesions in children and adults. Eur J Radiol 2013; 82: 837–849.

26.

Corti

Villafañe

Negroni

. Magnetic resonance imaging findings in AIDS patients with central nervous system cryptococcosis. Rev Iberoam Micol 2008; 25: 211–214.

27.

Hegde

Mohan

Lath

. Differential diagnosis for bilateral abnormalities of the basal ganglia and thalamus. Radiographics 2011; 31: 5–30.

28.

Lee

Park

Kwon

. Acute bilateral basal ganglia lesions in diabetic uraemia: diffusion-weighted MRI. Neuroradiology 2007; 49: 1009–1013.

29.

Sheerin

Pretorius

Briley

. Differential diagnosis of restricted diffusion confined to the cerebral cortex. Clin Radiol 2008; 63: 1245–1253.

30.

Ances

Sisti

Vaida

. Resting cerebral blood flow. A potential biomarker of the effects of HIV in the brain. Neurology 2009; 73: 702–708.